Modeling cascading failures of urban rail transit network based on passenger spatiotemporal heterogeneity

To explore and alleviate large-scale failures under urban rail transit network accidents, cascading failures are usually investigated integrating the network topology and passenger flow. However, the impacts of the heterogeneity of travel demand and its interactions with the spatiotemporal heterogeneity of rail passenger flow on cascading failures are still left unknown. In this study, the dynamic changes of origin-destination travel demand under failures, the time-delay of cascading failures and the interaction between these dynamic properties and spatiotemporal heterogeneity of passenger flow are modeled and explored in an improved disaster spread theory model. The approach is applied to Shanghai urban rail transit network under different attack scenarios. Results show that the scale of cascading failures on Shanghai urban rail transit network importantly depends on the ratio of passengers changing travel demand and time-delay of cascading failures. Besides, large in-flow of stations adjacent to the failed station would not only cause fast spread of cascading failures at the early stage but also result in many failed stations in the end. The heterogeneity of passenger flows in different directions on the network would reduce the scales of cascading failures. It is also found out that the circle line could obviously prevent the spread of cascading failures.